/Menu-Updater

The weekly dining hall menu updater is a project designed to automatically update the Dining Hall Menu page on the Fresno State Housing website by fetching the current week's menu from the Fresno State University Dining Services page, parsing it into a JSON object, and then uploading it to the Housing website's database.

Primary LanguageJavaScriptMIT LicenseMIT

Guide to using 'Weekly Dining Hall Menu Updater' 🍴

Overview 📝

The Weekly Dining Hall Menu Updater is a project designed to automatically update the Dining Hall Menu page on the Fresno State Housing website by fetching the current week's menu from the Fresno State University Dining Services page, parsing it into a JSON object, and then uploading it to the Housing website's database. The project is divided into multiple components, including fetching the menu PDF, converting it to JSON using the ChatGPT API, and updating the website's database with the new menu data. The project is designed to be run as a cron job, but it can also be run manually. It is also designed to work specifically with the Fresno State Housing website and the Pods Framework, but it can be modified to work with other websites that use WordPress.org.

Table of Contents 📚

  1. Overview
  2. Table of Contents
  3. Before You Begin
  4. How to Use
  5. Fully Autonomizing the Process
  6. How it Works
  7. Future Plans
  8. Resources
  9. Credits
  10. Conclusion

Before You Begin 🚀

Before you begin using the Weekly Dining Hall Menu Updater, you will need to have or install the following your computer:

How to Use 📖

Follow the steps below to use the Weekly Dining Hall Menu Updater to update the Dining Hall Menu page:

  1. Clone the repository to your computer.

    • If you have Git installed, you can clone the repository by running the following command in your terminal or command prompt:
    gh clone https://github.com/Web-Jose/Menu-Updater.git
    • If you don't have Git installed, you can download the repository as a ZIP file by clicking the green Code button at the top of the repository page and then clicking Download ZIP. Once the ZIP file has finished downloading, extract the files to your computer.

  2. Set up Node.js and the required Node.js packages.

    • If you haven't already, install Node.js and the required Node.js packages by following the instructions in the Before You Begin section.

  3. Replace the OPENAI_API_KEY in the Menu_PDF_To_JSON.js file with your own API key.

  4. Run the Menu_PDF_To_JSON.js file using Node.js.

    • Open your terminal or command prompt and navigate to the project directory. Then run the following command:
    node Menu_PDF_To_JSON.js

  5. Upload the Menu.json file to a repository.

  6. Add the code from the functions.php file to the functions.php file of your website.

    • If you're using a WordPress.org website, you can find the functions.php file in the Theme File Editor.
    • If you are Updating the Menu on the Fresno State Housing website, you can skip this step.

  7. Change the value of the $url variable in the functions.php file, found in the Theme File Editor, to the URL of the Menu.json file.

    • If you're using GitHub, you can find the URL of the Menu.json file by clicking on the file in your repository and then clicking the Raw button.
    • Save the changes to the functions.php file.

  8. Use the WP Crontrol plugin to schedule the dining_menu_updater function to run every week.

    • If you want to update the Dining_Hall_menu pod as a cron job, you'll need to have the WP Crontrol plugin installed on your website. You can install the WP Crontrol plugin by following the instructions here: https://wordpress.org/plugins/wp-crontrol/
    • Once you have the WP Crontrol plugin installed, navigate to Tools > Cron Events in the WordPress dashboard.
      • Click the Add New button.
      • In the Event Type field, select Standard Cron Event.
      • In the Hook Name field, enter dining_menu_updater.
      • In the Next Run field, enter the date and time you want the function to run.
      • In the Recurrence field, enter weekly.
      • Click the Add Event button.

  9. Check the Dining Hall Menu page to make sure the menu has been updated.

    • If the menu has not been updated, check the Menu.json file to make sure it has been updated.
    • If the Menu.json file has not been updated, check the Menu_PDF_To_JSON.js file to make sure it is working properly.
    • If the Menu_PDF_To_JSON.js file is working properly, check the functions.php file to make sure it is working properly.
    • Ensure that the variables and field names used in the Menu_PDF_To_JSON.js file and the functions.php file match the variables and field names used in your website's database.

Fully Autonomizing the Process 🤖

If you want to fully autonomize the process, you can create and run a .bat file to automatically run the Menu_PDF_To_JSON.js file every week. To do this, follow the steps below:

Step 1: Create a .bat file

  1. Open your text editor or IDE and create a new file.

  2. Copy and paste the following code into the file:

    set OPENAI_API_KEY=[YOUR_API_KEY]
node pdftext.js

  3. Replace the YOUR_API_KEY in the previous code with your own API key.

  4. Save the file as Weekly_Dining_Hall_Menu_Updater.bat.

Step 2: Create a Task Schedule

  1. Open the Task Scheduler app on your computer.
  2. Click the Create Task button.
  3. In the General tab, enter a name for the task.
  4. In the General tab, select Run whether user is logged on or not.
  5. In the General tab, select Run with highest privileges.
  6. In the Triggers tab, click the New button.
  7. In the Triggers tab, select Weekly.
  8. In the Triggers tab, select the day and time you want the task to run.
  9. In the Triggers tab, select Repeat task every and enter 1 in the field.
  10. In the Triggers tab, select for a duration of and enter Indefinitely in the field.
  11. In the Actions tab, click the New button.
  12. In the Actions tab, select Start a program.
  13. In the Actions tab, click the Browse button and select the Weekly_Dining_Hall_Menu_Updater.bat file.
  14. Click the OK button.

Step 3: Create a GitHub Actions Workflow

  1. Create a new folder on your computer and name it .github.

  2. Create a new folder inside the .github folder and name it workflows.

  3. Create a new file inside the workflows folder and name it github-json-menu-updater.yml.

  4. Copy and paste the following code into the file:

    name: GitHub JSON Menu Updater

on:
push:
paths:
  - "menu.json"
jobs:
  commit-and-push:
  runs-on: ubuntu-latest
  steps:
    - name: Checkout repository
      uses: actions/checkout@v2

    - name: Setup Git User
      run: |
        git config user.name "GitHub Action"
        git config user.email "action@github.com"

    - name: Commit and Push Changes
      run: |
        git add menu.json
        git commit -m "Auto commit: Updated menu.json"
        git push origin main

  5. Replace the user.name and user.email in the previous code with your own name and email.

  6. Save the file.

How it Works 🛠️

This is a deep dive into the code of the Weekly Dining Hall Menu Updater. If you're not interested in the code, you can skip to the Future Plans section.

Menu_PDF_To_JSON.js

The Menu_PDF_To_JSON.js file is the main file of the Weekly Dining Hall Menu Updater. It is responsible for fetching the current week's menu from the Fresno State University Dining Services page, converting it to JSON using the ChatGPT API, and then updating the Menu.json file with the new menu data. The file is divided into multiple sections, each of which is responsible for a different part of the process.

Section 1: Importing the Required Node.js Packages

  • The first section of the file imports the required Node.js packages. The axios package is used to fetch the menu PDF from the Fresno State University Dining Services page. The pdf-parse package is used to convert the PDF to text. The openai package is used to convert the text to JSON using the ChatGPT API.

    • The axios package is imported using the following code:

      const axios = require("axios");

    • The pdf-parse package is imported using the following code:

      const PDFParser = require("pdf-parse");

    • The openai package is imported using the following code:

      const { Configuration, OpenAIApi } = require("openai");
const configuration = new Configuration({
  apiKey: process.env.OPENAI_API_KEY,
});
const openai = new OpenAIApi(configuration);

    • The OPENAI_API_KEY is stored in a .env file, which is not included in the repository. You can find your API key on the OpenAI platform.

Section 2: Fetching the Menu PDF and Extracting the text

  • The second section of the file fetches the menu PDF from the Fresno State University Dining Services page. This is done by using the current date to find the date of the past Sunday, which is the first day of the current week. The date of the past Sunday is then used to find the URL of the menu PDF as they follow the format of menu-(YYYY)-(MM)-(DD).pdf on the Fresno State University Dining Services page. The URL is then used to fetch the menu PDF using the axios package. Once the menu PDF has been fetched, the pdf-parse package is used to extract the text from the PDF.

    • The date of the past Sunday is found using the following code:

      function getPastSundayDate() {
  const today = new Date();
  const dayOfWeek = today.getDay();
  const diff = dayOfWeek === 0 ? 7 : dayOfWeek;
  const pastSunday = new Date(today.setDate(today.getDate() - diff));
  return pastSunday;
}

    • The date is formatted using the following code:

      function formatDate(date) {
  const year = date.getFullYear();
  const month = String(date.getMonth() + 1).padStart(2, "0");
  const day = String(date.getDate()).padStart(2, "0");
  return `${year}-${month}-${day}`;
}

    • The URL of the menu PDF is found using the following code:

      function generateLink() {
  const pastSunday = getPastSundayDate();
  const formattedDate = formatDate(pastSunday);
  const link =
    "https://auxiliary.fresnostate.edu/association/dining/documents/rdh_menus/menu-" +
    formattedDate +
    ".pdf";
  return link;
}

    • The menu PDF is fetched and the text is extracted using the following code:

      async function fetchAndExtractText() {
  try {
    const link = generateLink();
    const response = await axios.get(link, {
      responseType: "arraybuffer",
    });
    const data = response.data;
    const pdf = await PDFParser(data);
    let text = pdf.text.replace(/\s+/g, " "); // Replace multiple spaces with a single space
    text = text.replace(/(\w)([A-Z])/g, "$1 $2"); // Add space between words and capital letters
    text = text.replace(/(\d{4})(\d{1,2})/g, "$1 $2"); // Add space after the fourth digit for 5 or 6-digit numbers
    text = text.replace(/([a-zA-Z])(\d)/g, "$1 $2"); // Add space between letters and numbers
    return text;
  } catch (error) {
    console.error("Error fetching or parsing the PDF:", error);
    throw error;
  }
}

Section 3: Converting the Text to JSON

  • The third section of the file converts the text to JSON using the ChatGPT API. The ChatGPT API is a natural language processing model that can be used to convert text to JSON. The ChatGPT API is used to convert the text to JSON by using the openai package to send a request to the ChatGPT API. The request includes the text to be converted and the prompt to be used. The prompt is a string of text that is used to tell the ChatGPT API what to do with the text. The prompt used in this section is "Convert the following text to JSON:". The response from the ChatGPT API is then used to update the Menu.json file.

    • The prompt to include in the request is generated using the following code:

      async function GeneratePrompt() {
  try {
    const PDFtext = await fetchAndExtractText();
    const prompt = PDFtext;
    return prompt;
  } catch (error) {
    console.error("Error generating the prompt:", error);
    throw error;
  }
}

    • The request is sent to the ChatGPT API using the following code:

      async function createChatCompletion() {
  try {
    const prompt = await GeneratePrompt(); // Wait for the prompt text to be generated
    const completion = await openai.createChatCompletion({
      model: "gpt-3.5-turbo",
      messages: [
        {
          role: "system",
          content:
            "You are given a list of meals for each day of the week. Your task is to convert this information into a JSON format following a specific structure. Each day of the week will be represented as a key in the JSON object, using the word form (e.g., Sunday, Monday, Tuesday, etc.). For each day, there will be four sub-keys: Breakfast, Lunch, Dinner, and Dessert. The value corresponding to each sub-key will be an array containing individual dishes served during that meal. Each dish should be represented as a dictionary, where the type of dish (e.g., Entrée, Side 1, Side 2, Protein, etc.) will serve as the key, and the actual dish's name will be the value. Please note the following guidelines: If a meal includes multiple dishes of the same type (e.g., two different side dishes), use numbers to differentiate them (e.g., Side 1, Side 2). Before creating the JSON representation, please ensure to perform a spell check on the provided information to correct any grammatical mistakes or misspellings. Your JSON output should adhere to this structure for every day of the week, including all meal types.",
        },
        {
          role: "user",
          content: prompt,
        },
      ],
      temperature: 0.2,
      max_tokens: 2048,
    });

    • The response from the ChatGPT API is used to update the Menu.json file using the following code:

      // Extract the output message from the API response
const outputMessage = completion.data.choices[0].message;

// Convert the JSON-formatted string to a JavaScript object
const menuData = JSON.parse(outputMessage.content);

// Write the data to the Menu.json file in the same folder
fs.writeFile("Menu.json", JSON.stringify(menuData, null, 2), (err) => {
  if (err) {
    console.error("Error writing to Menu.json:", err);
  } else {
    console.log("Output saved to Menu.json");
  }
});

    • This closes the createChatCompletion function

        } catch (error) {
    console.error("Error creating chat completion:", error);
  }
}

    • To run the createChatCompletion function, use the following code:

      createChatCompletion();

functions.php

The functions.php file is responsible for updating the Pods database of the Fresno State Housing website with the new menu data. The file is divided into multiple sections, each of which is responsible for a different part of the process.

Section 1: How the WP Crontrol can call the function

  • The first section of the file is responsible for creating the dining_menu_updater action. This function is called by the WP Crontrol plugin to update the database of the Fresno State Housing website. The function is called by the WP Crontrol plugin using the following code:

    add_action("dining_menu_updater", "update_pods_from_github");

Section 2: Getting the Pod, checking if it exists, and setting values to 'None'

  • The second section of the file is responsible for getting the Dining_Hall_menu pod from the database, checking if it exists, and setting the values of the fields to None. The Dining_Hall_menu pod is the pod that stores the menu data on the Fresno State Housing website.

    • The function gets the Dining_Hall_menu pod from the database using the following code:

      function update_pods_from_github() {
    //Get Pod to make values None
    $InitialPod = pods('dining_hall_menu');

    // Check if pod exists
    if ($InitialPod->exists()) {
        // Get all fields.
        $fields = $InitialPod->fields();
        // Set all values to None
        foreach ($fields as $field) {
            $InitialPod->save([$field['name'] => 'None']);
        }
    } else {
        error_log('Pod does not exist: ' . $InitialPod);
        return;
    }

    • Gets RAW JSON data from GitHub using the following code:

      $url = 'https://raw.githubusercontent.com/Web-Jose/Menu-Updater/main/Menu.json';
$response = wp_remote_get($url);
// Check for errors
if (is_wp_error($response)) {
    error_log('Error in URL request: ' . $response->get_error_message());
    return;
}
// Get the JSON from the body of the response
$json_data = wp_remote_retrieve_body($response);

    • Check JSON data for errors using the following code:

      // Check if the JSON data was retrieved
if (empty($json_data)) {
    error_log('No JSON data retrieved from URL');
    return;
}
// Log the first 200 characters of the JSON data
error_log('First 200 chars of JSON data: ' . substr($json_data, 0, 200));

    • Convert JSON data to an associative array using the following code:

      $menu_data = json_decode($json_data, true);

    • Check if the JSON data was converted to an associative array using the following code:

      // Check the JSON error
switch (json_last_error()) {
case JSON_ERROR_NONE:
    error_log('No JSON decode errors');
    break;
case JSON_ERROR_DEPTH:
    error_log('JSON decode error: Maximum stack depth exceeded');
    break;
case JSON_ERROR_STATE_MISMATCH:
    error_log('JSON decode error: Underflow or the modes mismatch');
    break;
case JSON_ERROR_CTRL_CHAR:
    error_log('JSON decode error: Unexpected control character found');
    break;
case JSON_ERROR_SYNTAX:
    error_log('JSON decode error: Syntax error, malformed JSON');
    break;
case JSON_ERROR_UTF8:
    error_log('JSON decode error: Malformed UTF-8 characters, possibly incorrectly encoded');
    break;
default:
    error_log('JSON decode error: Unknown error');
    break;
    }

// Check for errors decoding the JSON
if (is_null($menu_data)) {
    error_log('Error decoding JSON from API');
    return;
}

    • Update the database with the new menu data using the following code:

      // Loop through each day in the menu data (e.g., Sunday, Monday, Tuesday)
foreach ($menu_data as $day => $meals) {
    // Loop through each meal type (e.g., Breakfast, Lunch, Dinner)
    foreach ($meals as $meal_type => $items) {
        if (is_array($items)) {
            //Loop through each food type (e.g., Entrees, Sides, Desserts)
            foreach ($items as $food_item) {
                if (is_array($food_item)) {
    // Loop Through each food item (e.g., Chicken, Rice, Beans)
    foreach ($food_item as $food_type => $item) {
                    // Get the Dining Hall Menu Pod
                    $DiningPod = pods('dining_hall_menu');
                    // Check if pod exists
                    if ($DiningPod->exists()) {
                        //Creating the field name
                        $field_name = strtolower($day . 's_' . $meal_type . '_' . str_replace(" ", "_", str_replace("é", "e", $food_type)));
                        // Save the value to the field name
                        $FieldValue = array(
                            $field_name => $item
                        );
                        // Update the pod
                        $DiningPod->save($FieldValue);
                    } else {
                        error_log('Pod does not exist: ' . $DiningPod);
                        continue;
                    }
                }
                } else {
        $DiningPod = pods('dining_hall_menu');
        if ($DiningPod->exists()) {
         $field_name = strtolower($day . 's_' . $meal_type . '_' . str_replace(" ", "_", str_replace("é", "e", array_keys($items)[0])));
         $FieldValue = array(
          $field_name => $food_item
         );
     $DiningPod->save($FieldValue);
                } else {
                    error_log('Pod does not exist: ' . $DiningPod);
                }
       }
      }
    } else {
        error_log('$items is not an array on day ' . $day . ', meal_type ' . $meal_type);
        }
    }
}
}

Future Plans 📅

  • Eliminate the need for Github Actions by adding git commands to the .bat file.

Resources 📦

Credits 🎉

  • Jose Cortes - Project Creator, Developer, and Maintainer.

License 📜

This project is licensed under the MIT License - see the LICENSE file for details.

Conclusion 🎬

Thank you for reading this guide. I hope you found it helpful. If you have any questions, comments, or suggestions, please feel free to contact me at JoseHCortes02@gmail.com.